Vanished Vikings of the West: Demise of the Western Settlement

In Part I, we looked at the Viking colonization of Greenland, and the failure of their settlement in America. In Part II we saw the fall of the Eastern Settlement. Now we will look at an even more mysterious disappearance, that of the Western Settlement.

Of the two Viking settlements, the Western Settlement was both smaller and more marginal. With some 90 farms, it was a quarter the size of the Eastern Settlement and may have housed 1000 people at its peak, possibly only half that. It was much further north and had a two-month shorter growing season and a longer winter. The settlers farmed, but at much lower intensity than in the Eastern Settlement. There were advantages, however. The location was less susceptible to Atlantic storms that could reach the southern tip of Greenland, it had much less drift ice, and was much closer to the walrus hunting region in Disko Bay. A large harbour seal population was nearby, and the lack of drift ice meant this remained reliable until the 14th century. But it was isolated and visits even from the Eastern Settlement became infrequent. Traders from Norway also would not normally come this far.

Initially, there was good contact between the two settlements. The Western Settlement was involved with the Newfoundland exploration. But very little is known about the Western Settlement during later years. The community was doing well around 1300 when the last of the churches there was build. In 1327, taxes were paid in ivory, suggesting the annual hunts were still taking place. After that there was silence.

An excavation of one farm, V54 at Nipaatsoq, showed that the house was a complex of connected rooms, where sheep were kept in a room in the middle. The farm had burned down around 1225 but was rebuild with a stone facade, a sign of some wealth. Unexpectedly, a small silver shield (less than 2 cm in size) was found that was made locally but which has a pattern that is derived from the Scottish Campbell clan and was first used in 1296. The same pattern was carved on a piece of bone and on a soap stone. The human story behind this shield will never be known, but it points at some level of contact with a distant world.

An important investigation was done at the so-called ‘farm under the sand’ (‘GUS’ on maps). The remains of the farm were discovered under a thick layer of sand and gravel, along the bank of a small river. The small farm had been used for a long period of time, 300 years: it may have dated from the very first settlement. But it was not inhabited continuously. There were at least two periods when it was unoccupied and only roaming animals entered. In both cases the farm was subsequently rebuild and re-occupied. The times when it stood empty may have to do with overgrazing, or it may be normal family politics where children marry and move on, leaving no one to take over the parental firm. The farm was abandoned for the final time sometime after 1310. The end was caused by the small river: the site was overrun by river sand, presumably after flooding. (When the farm was found, it was covered in 1.5 meter of sand and gravel. However, this was deposited a few hundred years later, when the nearby glacier had expanded.) After every abandonment, including the final one, sheep kept visiting the now empty building. It was part of a normal cycle, with as only difference that after the last abandonment, no one ever returned. One of the rooms had been emptied and locked. The valuables had been taken: this was an organized evacuation, not a disaster.

Excavating the farm under the sand

Several other farms have been excavated. They show that life in the settlement went on as normal until the end. The farms were well build and well equipped. A bath house was uncovered on one farm, with a wooden floor made from ship’s timbers. A die made from walrus tusk gives some indication how the long winter months were spend. There was no indication of how and why these farms had been abandoned. However, in these farms some highly precious objects were found, such as iron thongs and looms. These would not have been willingly left behind by the owners; and this was not just in one farm but commonly. It seems likely that the owners had died rather than moved on. The farms had been left to decay; no one ever came back, not even for the wood which was perhaps the most valuable item. There was no one left who needed it.

They disappeared in silence. The last written account of the Western Settlement came from Ivar Bardarson, a priest who had been send to Greenland by the bishop in Bergen in 1342. A bit ironic, he was send mainly as a tax collector. At the time, taxes were due both to the government and to the church. The Greenland Vikings had last paid church taxes in 1327, as a shipment of ivory, but after that no more payments had come. Bardarson was send as a business manager, with the task to implement a parish system. He obviously send a positive message back, because the next year a trader went to Greenland and returned ‘richly laden with goods’, and 1346 and 1347 saw several further trade visits, as well as the very unusual occurrence of a small and not really ocean-worthy Greenland ship reaching Iceland. But all contact stopped in 1349 due to the Black Death in Europe. (The long journey kept Greenland safe from the plague: any ship carrying the disease would have succumbed long before arrival.) Bardarson was finally able to return to Norway in 1364.

His report is entitled “The Description of Greenland according to Ívar Bárðarson”, and was originally written around 1360. We do not have that original text: the existing version was compiled in Bergen from several other manuscripts, soon after 1500. We do not know how much of the text is original. For instance, there is a sentence on sailing directions from Iceland to Greenland, and the fact that the old route has become difficult due to the presence of sea ice, which may have been inserted by others. But most of the document is a listing of farms and this is thought to be reliable.

There is only a brief mention of his visit to the Western Settlement. He was asked by the leader of the Eastern Settlement to go there because of reports that the ‘skraelings’ (Inuit) had taken over the site. He took several ships and an armed party, and sailed up the Ameralik and Lysufjord fjord to the largest farm where he investigated six farms, four small and two large. His findings made it clear that no taxes should be expected from it:

In the Western Settlement stands a large church, named Stensnes [Sandnes] Church. That church was for a time the cathedral and bishop’s seat. Now the Skrœlings have destroyed all of the Western Settlement; there are left some horses, goats, cattle, and sheep, all feral, and no people either Christian or heathen.

It mentions that they went on shore to kill the roaming animals (for meat, presumably). Bardarson concluded that the settlement had failed completely after an attack by the Inuit. The details are dubious. The style of writing is clearly different and sketchy compared to the rest of his writing and seems to have been written by someone else. Cattle and horses would not have survived the winter and feral cattle are therefore implausible. Sheep and goats are hardier. It seems fair to assume he went and found it to be deserted, but details such as the feral cattle seem dubious. The destruction that is mentioned is also not supported by the archaeology which only found abandonment: there is no indication for any outside attack. The failure lay within.

The report does suggest a fairly sudden failure, where the Eastern Settlement did not know what had happened. Most likely, the disappearance of the people was discovered during the annual walrus hunt, and Bardarson’s trip was the next summer. Bardarson would have planned a visit anyway as the Western Settlement was a quarter of this task. It would likely have taken place after he had finished his main work in the Eastern Settlement and must have been brought forward. We can guess that it was in the late 1340’s or early 1350’s. It cannot have been later than 1363, or earlier than 1343.

Excavation of a farm in the Vathnaverfi region of the Eastern Settlement, showing the typical construction of rooms. There would have been turf roofs. Photo taken in 1939

It appeared the settlement had failed within a few years, perhaps even within a single year. The dead had been buried but we do not know where or by whom.

Archaeology has moved on since the days of excavating bricks and ornaments. The possible solution to the mystery of the Western Settlement came from an unexpected angle: flies.

When the Vikings came, they brought their insects with them as unwanted cargo. Some lived in the houses, some on the meat, some in the manure. The Vikings flies liked warmth and when the Vikings left, their flies went extinct too. The flies were as domesticated as the cows. When the Vikings brought animals into their houses for the winter, the flies would have had a field day. At the start of summer, when the animals were thrown out, the Vikings would shovel the content of the house out as well (pretty smelly, one can imagine, after 7 months of dung production), and start afresh. The floor debris found in the houses therefore dates from the last year of habitation, and this includes the remains of the flies.

The flies at the farm at Nipaatsoq showed that conditions had broken down. There is one particular fly which breeds in meadow grass: its remains are commonly found in the farms. But this fly disappeared in the top layer. Evidently, the fodder for the final winter was of poor quality, and contained little grass. There was an abundance of flies that feed on warm (and human) faeces. These flies were not unusual in the farms but never in these numbers. At the very top of the layer, the warm-weather flies disappear and the local cold-temperature flies take over. Carrion flies appear in the bedroom. Carrion flies are common in Inuit settlements but not in the Viking farms: they left no meat to go spare. What did these carrion flies eat?

There is an obvious clue to what the human occupants ate, in that final winter. The butchered remains of five cattle were found in the house. They had eaten their herd. The scattered and butchered remains of two large hunting dogs showed that they met the same fate. The bones were found in the larder, the hall and the bedroom: the people had stopped taking care of the house. The other bones that were found were from hares and ptarmigan. These provide meat but no fat, and a person cannot survive on them in arctic conditions. It is called ‘rabbit starvation’.

Did the occupants survive the winter? We do not know. The remains of a calf and lamb shows that they lasted at least until early spring. Did they die in the bedroom? No human remains were found: perhaps someone buried them but who or when we do not know.

Excavation of other farms repeated the story. At Tummeralik (a mid-sized farm) five dogs had been killed and eaten; in the much larger farm at Sandnes, nine dogs met this fate. At the fields of the farm beneath the sand, DNA analysis in a soil core showed that cattle disappeared within at most a few years. Either they had been killed and eaten, or they did not survive without human attendance.

It is a stark tale. This was no organised evacuation: it was hopeless starvation. It paints a picture of people trapped inside for the winter, with insufficient food to see them through. Butchering of dogs has been known in Iceland as well. When food ran low, farmers would kill the dogs rather than running out of meat to feed them. But in Iceland, new dogs could be bought. In Greenland, the loss of their dogs would have meant that after the winter there would be no hunting, and no meat. It was desperation.

The event must have taken at least two years. There had been a bad summer without hay, most likely because of snow cover in the prime growing season. There was insufficient hay harvested to feed the cattle during winter. The meat ran out, indicating that the caribou and/or seal hunts had failed as well. The inhabitants faced a desperate winter in which they ate every living animal in the house, and finally ran out of food. When the brief summer came, the survivors buried the dead, but there were not enough of them to rebuild.

When did this happen? Carbon dating indicates it happened around the mid 1300’s although with a large uncertainty. The loss of cattle around the farm beneath the sand is dated to 1352+-36 years. The description by Ivar Bardarson is consistent with this: the disaster struck somewhere around 1350, and was so fast that by the time the hunters of the Eastern Settlement noticed their absence, none were left alive.

Great Hall, Western Settlement. Source

Part of the answer to the mystery of the Western Settlement appeared in April 1982. This was when El Chichon blew up, perhaps the most underrated eruption of the 20th century. El Chichon was ‘only’ a VEI 5 but it compensated by a massive sulfur amount, and it managed to explosively put this into the stratosphere. Its climate effects were countered by the strong El Nino of 1983 and this is one of the reasons that the eruption is under-appreciated. The world still cooled in 1983 and 1984, but only by 0.2 to 0.3 C. Wind patterns played a role: Europe warmed from the stronger southwesterlies, while Greenland cooled.

When the impact was modelled shortly after the eruption, it predicted that the volcanic cooling would be strongest in the far north, and especially in the spring of 1984. This 2-year delay is caused by local effects. The cooling in the far north regions comes from snow which reflect more sunlight, and from sea ice which stops the sea from warming the air. In 1983, the winter snows would have stayed on the ground longer, and this would allow sea ice to form earlier and more extensively, with its effect in the spring and summer of 1984.

The predictions were shown to be correct: the winter of 1983/84 in Greenland was exceptionally cold. And this was especially so in the region of the Western Settlement, where January temperatures were down by 10C compared to normal, and spring was 2C colder than usual. July was normal, but August again was 2C too cold. The two arms of the fjord along which the main farms were build were blocked by a patch of dense pack ice and some lighter ice into June. Normally, these fjords are navigable in April. In early June 1984, the only snow-free area in the southern half of the Western Settlement was the small valley surrounding the Sandnes manor. By late June the ice had gone but little grass had grown.

If this had happened in the time of the Western Settlement, the ice would have made it very difficult to reach the seal areas at the head of the fjord in time for the spring hunt, while late grass growth would have reduced the amount and quality of the hay.

The Pinatubo eruption of 1991 also led to two cold winters in western Greenland, with temperatures of 2 to 3 C below average. The area around the Western Settlement had the largest change, and seems to be the most sensitive. This is because the sea here is normally ice-free, but a very cold winter can bring in the ice. That gives a much larger change than in an area which normally already had sea ice.

What would the impact have been? Models for farming yields have been run, which estimate fodder consumption and domestic consumption in the Western Settlement. They show that their way of life was remarkably resilient against bad years. Their mixed farming/hunting strategy provided insurance against failures. A 30% drop in hay production could easily be accommodated. Even a 70% drop was survivable by reducing the size of the herd and increasing the hunt. A recovery period would be required in order for the herd to rebuild. However, if several very bad years followed each other in close succession, the herds could drop below a critical level. And if the next summer was delayed, the herd would need fodder for longer and this could cause disaster.

Human consumption could also become problematic. Hunting opportunities were limited in the early spring. Only the ringed seals could have made a difference but the Vikings did not hunt these, for lack of harpoons. The hunt for harbour seals and harp seals in June was crucial in bad years, as it provided fat when most needed. The settlement appears to have failed in May or June, as indicated by the presence of a calf and lamb and the fact that the cattle manure had not been cleared. If ice had blocked access to the sea and perhaps for the same reason the harbour seals had not appeared, after a few bad years this could have doomed the settlement.

What do we know about the climate in these years? The plot shows North American tree rings. It shows the longer term trend and the cold years of the 1450’s discussed in the previous post. But is also shows a significant decline in the 1340’s. Something had gone wrong with the world around 1345.

Esper et al, 2002, Science, Vol. 295, Issue 5563, pp. 2250-2253

Ice cores have confirmed this. They show two large eruptions in the 14th century: one (possibly a double) eruption occurred around 1330 in the northern hemisphere, and a tropical eruption hit around 1345. The second eruption was the largest, and this is our main suspect. The same eruption has also been blamed for the Black Death. It caused very wet summers in Europe, and just as had happened in 541 AD, this allowed the rats and their fleas to spread. From 1347 onward, the Black Death devastated Europe and closed the trade with Greenland. And at the same time, and from the same volcano, the Western Settlement was dying.

For both the Eastern Settlement and the Western Settlement, the end coincided with a significant volcanic eruption. We cannot prove these caused their demise. But the eruptions would have had an impact. The settlements could cope with a bad year. They could also cope with the slow cooling from the Little Ice Age, which allowed them time to adjust their farming methods. But a major eruption can cause several bad years in succession. That tested their resilience to the limit. And perhaps already suffering from lack of trade, falling population and depletion of resources, it pushed them beyond those limits.

There are lessons to be learned from the failure of the distant and isolated Viking settlements on Greenland and Newfoundland. We are not an island: to build a new world, sometimes we do need help from the old world. But we should not forget their human story, of normal people trying to run their lives in such a beautiful but forbidding place. For many years they succeeded, against all odds. In the end, the odds became too much; the nation they build passed on to others. But we should remember them, in their successes, their adventures, their failures, but above all in their humanity. They too were people like us.

Albert, June 2021

The first steps on a journey of legend. The settlers arrive

303 thoughts on “Vanished Vikings of the West: Demise of the Western Settlement

  1. Kilauea still inflating fast, another DI event could have just started but the trend is overall still up and I cant see that changing without something major, be it a rift intrusion or an eruption somewhere.

    GPS has exceeded the point it reached in December today, and with the overall trajectory of the past week will pass it totally within days, which could be the timeline in question regarding eruption… If there isnt an eruption within this week I think we should set our expectations a bit higher, could be an impressive show 🙂

    • It might require some additional pressure to break through Halema’uma’u Crater than last time. There is a significant added weight from the lava lake after all, and the previous pathway created in December seems to have sealed up or it would have started leaking already. When it erupts there will be a pressure excess to be released at the start. If the eruption does happen at Halema’uma’u I bet there will be some impressive foundering and dome fountains of lava. I wish there was a live webcam like those of Iceland.

      • There is a curious deformation of the ring fault though, CRIM is rising but OUTL is falling, like there is shallow magma there. I would not be surprised if there is a ring fault eruption, which is good for those of us that want lava because eruptions here are like mini flood basalts, like the 5 million m3 erupted in an hour in December 1974… 🙂

        The south caldera has not seen a lot of lava for centuries, it could be that its luck has run out…

    • I presume that if a new pathway needs to be established, then there should be quite a few eqs to follow its development.

      • New pathway yes, earthquakes yes, long time… no

        December eruption was preceded by less than an hour, and no felt quakes. People saw the beginning of the eruption and said it just happened like that. The fact this is the most well monitored volcano on earth taking one of the best teams of volcanologists by surprise shows how sudden this was. It has always been this way too, historical accounts of summit eruptions actually being preceded by anything is restricted to 1959 which was not in Halemaumau.

  2. This is the first part of my previous video (where the battery was running out). I think I may have got too close where it lost focus (lol)

    • We were right to say that the cornice couldn’t hang on for much longer.

    • Wow! Doug…. these are great… You took the views that i’d like to see…. the back side and an over view of the older brother which is almost taken over and a grand over view of the whole flow. i’m blathering but anyway… i really enjoyed this …………. Thanks…..

    • Absolutely fantastic, thanks for posting that. Really shows the scale of the ‘holes’ where the lava goes.

    • Thanks for this. The best overall footage of the summit and exit of the flow available on the internet. I’ve been hoping for something like this to illustrate how the lava tube is being fed.

  3. I presume the lava is heading for Meradalir since the channel is filling up again. I have noticed a lava glow emerging at 22.25 pm next to the vent overflow near Bob RIP. This is on the original Geldingadalir webcam.
    Do we have overturning, or glow from the failing light conditions, or perhaps a broken lava tube?

  4. Ever since the flow into Natthangi stopped, there have been between 1 and 3 million cubic meters flowing into Meradalir. But the only glimpse into it was 67dougs panning shot, for about half a second, in the end of the video from a distance.

    It seems like Meradalir has been filled all the way to its end, and is forming a lake; but the lake at the end seems to be less high than what one would expect with a few million cubic meters. So either it is already flowing out of Meradalir by the exit, or, more likely, is is stacking up in the upper parts of the valley, creating a larger gradient than what a lake alone would support. That upper part of the valley is also where most of the gas/fumes/steam seems to come from, most of the time.

    Someone needs to make a proper overflight and see how far the lava has made it to the northeast…

    • Yes. We’re blind in that direction now, and the southern part of Geldingadair (although the general public will be videoing this area). Gutn Tog has done several good video clips from Meradalir; but he is working this week and so doesn’t have capacity to visit the eruption site.

      I noticed that he has started to put his email address on his videos; perhaps he could be invited to pay a visit to the VC by a friendly Dragon?

    • It goes in phases. There will be a week or more where the lava goes north, then something will break and it goes south. The slope is steeper to the north so most of the time it goes there. I expect Natthagi to remain as it is until the next break. The road should be safe for now.

      • There was an interesting slide shown at the AGU webinar (I think you were there?) showing predicted lava coverage if the current eruption rate is maintained; it didn’t show the road being covered until 2023 or 2024.

        • I think that was based on the lava first filling up Meradalir. If that route blocks for any reason things will evolve faster. But the Meradalir direction is the preferred one based on the topography

    • If you’re speaking of that area where the left stream drops out of the crater, I agree.

      Spurting and flowing. Game changer me thinks!


      • yes, that area to the left. I posted a couple hours earlier or so about this leak.

        • Seems to be fighting the sloughing off from the walls around it.

          Hard to kinda figure out how much lower the leak entrance is compared to the right flow stream.

          As long as it doesn’t close off, it will make things interesting. 67doug’s great video show the opposite crater walls kinda on the thin side, so whatever pressure the leak can take off that will be great!


    • Could this possibly be related to the former fissure located here? Remember how Rag-Nar was two vents at first, and the one closest to Nordri-Sudri ‘migrated’ to take over the other.

  5. Big gush at 00:20:25 am. If anything, it looks like the flow rate has increased slightly.

    • And another almost identical big gusher at 00:28:37 am. If we thought that the volcano was quieting down, this thought has been swept away by the last 2 gushers.

  6. Right, so all my speculation earlier was rubbish. That’s fine by me. Looking at the webcams I can see the pulsing is still present, with some voluminous outpourings, too. But the conduit must be very open now.

    Nonetheless, I am going to stick my neck out and speculate we may yet be in for a change. I am slightly bothered by activities around the Keilir hill area where we had small earthquakes around 2-4km deep. Our conduit changed after those. Meanwhile the KRI drumplot charts show some interesting, large wet tornillos.

    Something is going on down there….(cue Jaws theme music).

  7. Things have looked very stable for the last hour at least.
    There’s persistent bubbling just below where the left overflow channel is, so it looks like lava may be coming through a crack in the cone.

  8. According to tremor graphs it seems that the volcano does not finished the pulsing phase….

  9. Thickness of the lava layers at June 2nd (maybe posted before):


    • Enjoyed the video thanks! With people watching on,it gives you an idea how big the cone and flow is.Far cry from 19th March when it all started.

      • So much lava flow and no lava flows visible to the valley… where all lava is going?

    • Image foreshortening changes the perspective which is why the earlier share overflight video from 67Doug was significant.

  10. Interesting trilogy. In this part the history of (missing flies) fascinated me and also the knowledge about El Chichon having an effect in the North while ENSO was protecting the rest of Europe.

  11. I am staying in a region that I hated most of my life, weekend suburb for ppl from Munich, summer resort for Germans, Dutch and few British, Lake Garda and Monte Baldo.
    Since keeping busy with plate tectonics including volcanism and fauna&flora I know how special this is.

    The rocks were carved out between 70 and 230 mya in the Triassic Tethys, then pushed north, and the lake is the result of four quarternary glaciations, most important the Wuerm glaciation.

    Wonderful plants, staying up in the mountains near Tremosine with a view best place. Airport for Britons: Verona (from Gatwick). Mediterranean climate, beautiful, too many Fritz though. No volcanism ever, sorry.

  12. First of all I want to give my thanks to these 3 blogs about the Greenland Vikings. A fascinating story which Albert compellingly tells.
    I sometimes wonder if Albert shouldn’t be one of those ‘personalities’ that explains scientific subjects to the layman.
    You’re a good storyteller and your knowledge seems to have no bounds.

    A timelapse from the close up camera this time.

    Yesterday 10 june 18h30 until 09h00 this morning.
    After a brief seemingly quiet period in the beginning, it gets in overflow mode this time.
    It’s fascinating to see where the lava goes and doesn’t go.

    For those who don’t know it yet, started creating timelapses from all of their cameras recently.
    So if you are looking for timelapses from the Natthagi camera, which I haven’t uploaded (I recorded a lot but missed some of the main events, that’s why) take a look at their youtube channel

    • There do seem to be intermittent connections between the sinks and the emergent points, it’s very unpredictable.
      It’s also much easier to see Randall’s leak in action.


    13.21 pm, and on the Langihryggur camera which is zoomed in, I’m watching people walking across the lava from the cut off Viewing Hill back to the other side. I’d glanced at Tokolosh’s ‘3 webcams and a map’ stream, which is up on the tv, and had assumed the people I could see were from helicopter flights, but there’s no helis to be seen.

    • The camera zooms in at 13.08.26 to focus on people on the hill.

    • At 13.47 as well…

      I guess they have no issues with dying? Impossible to say whether it is safe or not…

    • It’s not such a biggie, I’ve found myself doing the same at Kilauea by accident.

      Was climbing a lava flow on the pali… thought ‘it’s getting a bit warm all of a sudden’… then noticed the lava looked *very* fresh… then ‘it’s getting bloody warm now!’. But I climbed it without incident and camped in a kipuka higher up, in a safe spot.

      • Are you sure you are the right person to warn New Zealand about geohazards?

      • There was that guy who lived in Royal Gardens up until 2012 too, he had to do a weekly hike to Kalapana that often involved walking over an active tube, especially 2008-2011, and also in the early 90s, before finally the lava took the place in 2012.

        None of that kipuka exists anymore, was taken out only a few months before the eruption of 2018.

  14. @Twisted One: your comments are appearing as a reply to Mike Ross’ celebrated announcement on comment page 1. I have no idea why this is happening.

    • If you go up little i post a youtube clip of it earlyer. Better quality on video 🙂

    • He/she provided a sense of scale though, which has been difficult to judge.

      • Live i wouldn´t notice him.

        That position must to be hot and smoky. With

    • Oddly, when watching the Theatre Hill video, I have often idly speculated how survivable climbing the vent could be. I imagined walking along the large back wall…etc. And then assumed I would perish in the lava.

      Never in my right mind would I have through some idiot would actually try.

    • What timestamp is this? the video won’t play for me for some reason.

      • He arrives on screen from the left (lower) around 11:05 on the screen time. He’s very small!

        • i couldn’t see him…. old eyes…. can barely find my glasses… but i can see the red arrow. i can’t believe he got that far… and one fall and he’s be all cut up. please forgive me for assuming it’s a man.

          • 1) It’s a man, guaranteed.
            2) The vent is massive and the changes is height of the lake is measured in 10’s of meters.
            3) Awesome ….

          • I think it is a little old lady trying to get close enough to see the volcano

          • Little old lady with impressive downhill sprint skills.

    • On my first trip to Iceland, we visited Geysir. While we were there, we couldn’t understand why one bloke kept looking at me. When we were walking back to the van, we realised why,I was wearing a green fleece and he must have thought I was a Ranger or something. As soon as we left, he decided to go under the ropes to get a closer look at Geysir, then he ran like hell as it erupted. He got away with it once and pushed his luck to do it a second time. I don’t think he was as lucky that time.

    • There was a small quake registered on the drum plot chart at @15:21 IS time.

  15. Thank you, Albert, for your remarkable, yet sad, story. Can’t stop thinking about it.

    • The dogs did it for me, it makes you think alright.
      Volcano Café Book Albert? Its just a thought.

    • Thank you. There should still be a sequel, on those volcanoes, but that will have to wait. I had been thinking about this story for a long time. Writing it down helped me. I am sorry if it was too hard!

      • Sometimes, life is hard, and not every enterprise is successful. History is, always, written by the survivors. Maybe, somewhere in those abandoned houses, lie household accounts, or chronicles of the leaders, waiting to be discovered…

      • i’d rather read a true story rather than a less accurate one that’s been prettified. Thanks, Albert.

  16. I am thinking the voluminous amounts of gasses being released must be proportional to a significant increase in lava volumes if the composition is the same and the conduit pathway is also the same.

    • It can also depend on atmospheric conditions at the time. Higher humidity leads to more visible vapours.

    • Up until yesterday there was a message saying that they had lost connection to the IMO cameras. That’s gone now, but the links go the the IMO homepage instead. They may have a small image of the last one taken with each camera though.

      • As far as I can see, they are refreshing again (check timestamps).

        • Yeah. Low quality but better than thumbnail in vedur.

          Two examples:

        • The original source for the webcam photos is no longer available. They had too much traffic to deal with. That’s why it’s now linked back to the IMO website. And the photos are smaller 🙁
          I attempted to gain additional access, but for the time being, this is the only sorceress I have access to.

          • Thank you for trying to get access. I emailed the IMO to ask for access too, but haven’t had a reply as of yet.

            Is the your site? If so, thank you.

            Is it possible to see more of the sequences, or is each photo only ‘live’ for the ten minutes that it is current?

          • Stars Die, yes it’s my site.
            It’s possible to see more, I intend to keep them for few days or weeks. At the moment I do not have way to display them in proper manner, but I may do so in the future.

          • Merlot.
            This is the only source I’m aware of right now, and it’s where I acquire the images.
            IMO only shows 8 images per camera. My intention is to show more, perhaps images from the last 24 hours.

          • For those who are interested I’ve recently included a slideshow with photos from IMO cameras. displaying the most recent 50 pictures.


            I hope it is enjoyable for someone. Unfortunately, I can no longer obtain larger pictures.

  17. Thanks much for your Greenland articles Albert.
    Very interesting read.

  18. Any word on the flow rate? It seems like there are some just monstrous pulses happening now.

  19. Currently, the rafting of the lava has multiplied and is virtually constant

  20. Earthquake swarm under Mount Hood has moved from the flank to under the peak, and shallower. Interesting!

      • Merlot: any way to view the time sequence of Meridalir valley ? that webpage you mention only seems to show the nagatthi valley (for me at least)

      • USGS hasn’t posted anything regarding ground deformation or gas emissions, so likely it’s just the magma chamber stretching in its sleep.

  21. It’s reported in ‘The Watchers’ blog of more increased sesimic activity at Michoacan region Mexico. Home of the famous Paricutin volcano- the one that erupted in a man’s corn field.Anyway there’s talk of a future new volcano- but must be said swams in the 21st century have come and go,so latest swam may lead to nothing. Been 2021 anything is possible-such the volcanic activity we’ve witnessed this year.

  22. From the 3D model, the lava flow is over the crest toward the old channel to Meradalir and should reach it in relatively soon. A few more weeks buildup in the current channel and the firehose might divert north. That will quickly reshape that area since the flow is triple the output of the northern cones back then.

    • If the cone fails it will come rushing down that hill like a portal to hell has been opened… 🙂

      Is lucky it will be pretty harmless, though it might also not be well observed which is unfortunate.

  23. Swarm of quakes around Kilauea summit today, and CRIM station has changed orientation. It might be showing some subsidence now but it is accelerating east and south rapidly. PUHI station is also showing an eastwards trend over the past days. Would seem that there is a rapid localised inflation within the 2018 caldera, which is pretty telling.

    All of this really only shows in the past week at most, same timescale as the abrupt increase in caldera spreading on the main page. It is similar to late last year but not exactly the same signal, there is less rift movement but the summit movement is a lot more noticeable, like it is a shallow source. I would expect the next few years to be one of mostly summit activity, until the deep pit of Halemaumau is filled in to at least the level of the wider caldera, around 5 years of hotspot supply.
    The rift conduit has to be fully open a long way down the rift for gravity to favor an eruption there instead, as in further east than Pu’u O’o, something that is not evident. The last time magma seems to have made it this far was around a year ago, but at present the conduit is probably only fully open to Mauna Ulu, maybe to Napau or Pu’u O’o at most. Probably this will not be a long term configuration, south flank movement is still strong, but the summit is too easy at the moment. This does fit into the timeline HVO proposes, rift eruptions again in maybe 5-10 years, but a bit different reason.

    • Correction swarm was actually yesterday, but still, quite a change. Maybe more curious is that there has only been one quake there today, and the expected DI event has already bottomed out, like it was cancelled early.

      HVO also thinks DI events are driven by magma density changes.

      • And now a couple of quakes around 6 km under the summit, including one of those small quakes that is below the usual detection range… The depth is interesting, at the base of the main magma chamber. It is also interesting that the quakes are in the north part of the caldera too, not just in the south.

        Also an earthquake on the rift zone down towards Pu’u O’o, and another small quake near Mauna Ulu that falls outside usual detection again.

        • DI event is deflation/inflation event. It is on the tiltmeter, where the signal has a sudden drop that is followed by a recovery to about where it began, though not always. Their cause has been mysterious but it seems the best case for it is that denser magma falls and when enough of it falls to a deeper level in the storage there is overall net loss up in the shallow system so the volcano deflates, while hotter magma rises so fills back rather fast. Denser magma can have many causes, possibly it is not as hot, or it is degassed, or both, or something else entirely. The changes though are minor, and it seems there is not actually a lot of correlation between DI events and at least summit eruptions. It may though indicate an exceptionally open system, with no obstruction to continuous flow, which would be expected for such an active place but is not easy to confirm.

          • I also find that alternative likely. I have thought about cumulates of olivine, that settle at the bottom of the magma chambers and could be the dense material that falls. Sills tend to have the shape of a saucer, not exactly horizontal but rather just barely dipping towards the centre.

            I find even more appealing the possibility that DI events happen when masses of semi molten rock detach from the roofs of sills or the platforms between sills and slide over the slippery magma-hot surface towards the central conduit of the volcano where they plunge into the magma shaft.

            To me this seems just natural because it is seen all the time in volcanic eruptions, when the incandescent walls of a volcanic crater start slumping towards the vent, or when large pieces of the lava flow crust or the cone itself are carried as lavabergs with the current.

            I also remember watching the Pu’u’o’o collapse of August 2011 when the floor of the crater, part solid, part melt, started flowing towards the centre. Halema’uma’u would sure have swallowed up all that rock!


    • CRIM seems to be moving away from the centre of the caldera, like UWEV and BYRL. The shallow Halema’uma’u magma chambers are inflating and pushing the three closest GPS away. It is perhaps still too early to know how good of a trend this is, however the tiltmeter shows very clearly that the amount of magma entering the summit is very high, this tilting is faster than it ever was in 2019, or 2020 until the December surge, I would say we are seeing a surge in supply.

      I wonder how long it will take for the ERZ to erupt, the lava lake at the summit stands almost as high as Napau, another summit eruption and it will stand at the same height. However the pressure at Napau is slightly less than at the summit, the connection is not perfect, I wouldn’t be surprised if the rift goes off this year, but I wouldn’t be surprised either if it does take 5 years, while 10 seems too much to me.

      • The movement started with a bit of inflation seen south of the caldera in late March. Now the movement is away from the caldera. It is so pronounced now that it is worth considering raising the warning level. That probably waits for an earthquake swarm.

        • That was the reasoning that HVO had last time for not raising it, the next swarm was an eruption… To me I dont think they should have ever lowered it, while Mauna Loa is really doing nothing particularly noteworthy abd is sitting at the same alert and has been for years now. It seems very missleading to even have Kilauea and Mauna Loa on the same alert at all really.
          Orange alert is an ongoing eruption, but unless there is no eruption for 90 days it is considered a pause, so technically still ‘ongoing’. I guess they have their reasons but this seems rushed.

    • Thanks for the heads up. No email from HVO on this yet.

  24. I am still wondering… What is the material that is creating ‘scabs’ on the newly erupted lava? It is funny, because lava is pretty much as hot as it gets when it erupts. Yet there are still ‘scabs’.

    • The ‘scabs’ form on the lake in the quiet periods between outbursts when there is little fountaining to disturb the surface.

    • Those are probably slabs of cooler lava that break off the inside walls of the crater.

    • They may form from low level spatter. Once you form the holy stuff it insulates very well. The surface eruption temperature is reported to be 1170C, by the way

  25. Love the smell of sulfur in the morning. Going to be a rain day where i am = no need to go to garden. So a big 500ml jug of swedish coffee and “pepparkakor” ginger cookies and plan for a day study little volcano… And nice read Albert. As always. Have nice weekend and stay safe. Don’t climb the cone. 🙂

  26. I would certainly like an Iceland sequel about that runny lava, where it comes from and how this might develop in the near future, by Albert, Carl, Hector, Chad or Jesper. Pics and films are pretty, but by now it changed a lot which leads to the necessity of some scientific background, also gladly some educated guess. Please!

  27. There’s a big steamy area just behind the cut-off flow on the Langihryggur camera, I wonder if some of the flow is resurfacing round the back?

  28. Yo volcano folks. On the close up camera feed of the Geldingadalur Volcano please scroll back to 10:53:40 this morning (camera time) and press play. Now, watch the sky right above the volcano. You’ll see a white glare move across the sky, left to right. It’s really cooking. If you zoom in on a screen grab, you’ll note it has a tail like a comet, but I think that’s a streaming artifact. By 10:44:00 or so it seems to turn around and head back right to left. When I first saw it I thought “Cool! Space Station!” Then I realized it was 10 AM there — also it turning around didn’t make sense for the ISS or a satellite. That’s just the initial impression I got, having observed the ISS a few times. I can’t identify it so for now, it’s a UFO in the strict, non-alien-beings definition of UFO. My guess it it’s a helicopter and we’re seeing the sun reflecting off its windows.

      • Some kind of small aircraft. I do in fact get the impression it’s a plane, now that you mention it. Just weird though. Happened the same morning there was a person literally trying to climb the volcano. That’s (the climber) some of the spookiest footage I’ve ever seen.

        • At what time stamp does the climber appear, please and thanks?

          The UFO is cool.

        • I was going to suggest the Starship Enterprise, but your explanation sounds a bit more rooted in reality.


  29. Is there any new update on composition or temperature or flow rate?
    This looks like the hell machine described a few months ago; increasing steadily.

  30. WIth respect to the comment (from Albert) “It goes in phases. There will be a week or more where the lava goes north, then something will break and it goes south. The slope is steeper to the north so most of the time it goes there. I expect Natthagi to remain as it is until the next break. The road should be safe for now.”

    I agree, I am just wondering how much of it is underground (potentially very far reach) vs. on the surface. The 3D model pretty much confirms our expectations I guess; a lot stuck to the slopes, but based on the image much of the flow is now below surface, and it did form a full lake by now, creating an almost level surface in the end of the valley where the flow is blocked. So now it can mostly flow underground all the way to the end of the valley…

    By the way, with respect to a couple of comments that came up earlier by some people that might read like describing a switch to “shield formation mode” at some time when the laval will start to overflow 360 degree like some of these chocolade fountains in candy stores: That is not really what happens in a shield volcano formation. The lava will always go predominantly in one direction for extended periods of time, weeks, months, maybe years, before switching over to another random direction. For a nice shield however, this averages out to the full 360 degrees over many switches over the course of decades to something roughly symmetric…

    Now, the sketchy part is that in principle with such hot low viscosity lava and high elevetion source, the volcano can create long-distance (i.e. shallow slope angle) outflows in any given direction where there’s no uphill slope, and hit something like the road in a relatively short amount of time just randomly. So that I guess is the main source of risk overall.

    Specifically for Natthagi, the valley floow is less than 0.5 km^2, and the basin is on average not much more than 10m deep. So without a barrier, given what happened in Meradalir, if the flow switches fully towards Natthagi for a few of weeks (at current flow rates), it would appear that that would be enough to reach the road, absent any barriers.

  31. Photo taken from flight this morning, looking towards the south. The crater is on the right side of the picture, the southern and northern parts of Merardalur in themiddle, and the eastern part on the left. The biggest current seems to be flowing there now. Photo / Sigurbjörg

    Lava flow from the eruption in Geldingadalir now seems to largely lead to Eastern Meradalur, in addition to raising the current lava flow in both Geldingadalir and Meradalir. However, there is less flow going down to Nátthagi. Then the magma current is cutting itself down in the crater and it is likely that continuous flow will replace the lava flows that have been there recently.

    This is what Þorvaldur Þórðarson, professor of volcanology at the University of Iceland, says. He went on a field trip yesterday and visited, among other places, Meradalir and the eastern part of the lava field. As can be seen in the attached photo above taken this morning, the lava has stretched further east (to the left of the photo) in recent days, but it is in the eastern part of Meradalir.

    “There was a lot of activity there. The edge of the pahoehoe lava field in Eastern Meradalir was all active and there was some movement in it, “says Þorvaldur. Conditions are now such that the lava has completely filled up the entire bottom of the Meradalir valleys and the lava field has begun to thicken considerably. Thus, the thickness of the lava is 10-12 meters in the valley, but 2-3 meters at the edges. “The Meradalir valleys are covered with lava,” says Þorvaldur. “This is the same all over the valley and it tells you that it’s all coming together.” Everything is heading towards a shield eruption.
    Þorvaldur has previously said that the signs are for the eruption to become a so-called shield eruption, but that a lava pond must first form over the crater and the flow must be in lava channels rather than on the surface. He says that everything still seems to be heading in that direction. The lava field can slowly and surely cut itself into the crater and build over the outflow. Then there will be continuous flow rather than spurts of flows. With a change like this, Þorvaldur says that there will be much less heat loss from the lava. He mentions as an example that a lava field that is in a closed channel and well insulated loses only 1°C per kilometer. In the case of an open channel, the heat loss is about 100°C per kilometer. Therefore, pahoehoe flows are usually formed when the magma is in a closed channel, but apal lava in open channels, as lava magma solidifies around 150°C.
    Asked whether it is likely that the lava flow will move entirely from Nátthaga, as most concerns have been that the lava could eventually go towards structures, Þorvaldur says that it is uncertain how the transport systems in the lava field will develop, but changes there can be somewhat faster.

    Today, a lot of magma seems to go in Geldingadalir, in compacting the lava up there, and then in the northern and eastern parts of Meradalir, but also a little further in the southern part of Meradalir and down to Nátthaga. “But it is clear that this is taking over the most,” he says of the eastern part.

    • Wow! Nice photo.

      A bit confused. Meradalir is filled and next valley is half filled. Normal, many days with the lava flowing there and no new videos and photos..

      • That’s also part of Meradalir, more specifically East-Meradalur. “Nameless Valley” is South-Meradalur and the main one is North-Meradalur

        • It may help English speakers if I mention that an Icelandic speaking friend confirmed that “dalir” is the plural of the singular “dalur” (as in fjorður/fjirðir). So Meradalir means “Mare Valley*s*”. Hence the fact that some of valleys we are discussing are, to English ears, nameless is not strictly correct because in fact they are denoted by the Icelandic plural.

    • Thanks, that is an excellent photo. Puts all into perspective, including how close to the coast and road everything is. And how there are no towns that close.

      Also, can everyone see a row of what look like spatter cones in line with the current rift and vent? Towards the bottom. They are probably eroded, but if this vent has erupted before might give an idea of how big the eruption can get.

      Equally on the left there is a taller cone that looks to have slightly fresher lava and is less eroded.

      For all this area is supposed to have erupted for a long time, it looks to have been quite active in the past.

      Some professional commentary on this would be great, maybe using the picture as a start of an article?

  32. Some observations on this article:

    I first noticed that no remains of animals or humans were apparently found within the houses. This would imply that no humans perished of starvation or cold inside their dwellings. I can imagine the last of survivors making one last desperate effort to reach the Eastern settlement once their situation became hopeless and they perished in the effort. I can see them making such an attempt if they determine that their current situation is completely hopeless if they do not and at least it would give some hope of rescue.

    Feral animals might be a real possibility depending on breed if the welfare check mission was the same year the settlement was abandoned. If the horses were of the Icelandic breed, they have a double coat and are hardy at much colder temperatures than standard horses. They are also much smaller and conserve body heat through less surface area. I am somewhat skeptical of feral cattle, though.

    • Thanks for this insight. There were remains of animals in the houses, including cattle. Cattle is the most meaningful as they could not have survived a Greenland winter. But indeed no remains of people were reported in those houses that were excavated. There was such a find in the Eastern settlement, in one of the outlying farms. So it seems either the people died elsewhere, or they were buried by the survivors. Because of the carrion flies, the latter seems more likely in that particular case. The presence of feral cattle (if true: I also have doubts) would show that at least one farm pulled through long enough to release the cattle. Gardar is most likely. because after El Chichon, that was the only area in the settlement found to be snow-free in June. But they too had eaten their dogs so were in deep trouble. Did the starving farmers go to Gardar for help? That seems possible.

      Without food, dogs or boats, the survivors could only have waited for rescue, not gone out themselves. I expect that they were waiting for the July-August walrus hunters of the Eastern settlement to drop in on the settlement. But by the time they came, it was too late. I guess that they would have come on the way back, towards the end of summer and this is when they reported finding no one. The survivors lasted until June but not much beyond that. In bad years, the western Vikings depended on seals to carry them through from June to August. This year, there were no seals or they were unable to reach or hunt them. They had no dogs for the seal (or caribou) hunt. Eating the dogs in even the best located of their farms showed how desperate the situation was, towards the end of winter.

      • The worst time in a famine is the spring, particularly May. I have learned this from watching things in places like North Korea. This is when the winter stores have given out and the earliest crops are still not yet ready (radishes in that case). May is generally the starvation month and is when one sees reports of things such as people eating grass.

        I recall having read some long time ago that the population had been in decline for some time with people choosing to leave the settlements in dribs and drabs over the years. Also, I suppose I didn’t catch the remains of the animals and for some reason thought it was talking about bones from animals that had been butchered. But the lack of cleaning of the house of the final year’s manure would tend to indicate an early spring abandonment.

        It certainly was a harsh and unforgiving existence.

        • Both butchered and died animals were found but it is difficult to find the details. The excavations were done decades ago, and nowadays the area has warmed a lot and remains have decayed. There are reports of a dead newborn calf. The slow decline is thought to have happened in the eastern settlement where one or more areas were abandoned and all valuables (wood) removed. Inuit stories talk about pirate raids (that would have been the British!) but no evidence of violence or destruction has been found. The end of the eastern settlement wasn’t as fast as that of the western one. I think it was a slow decline by attrition, as life expectancy declined and the walrus and seal hunt became more dangerous – helped by the fact they no longer had seaworthy boats. The people probably moved to the main port on the coast, but much of that farm including possible part of the cemetery has been lost to the sea. I don’t think they were rescued. There would have been a record or stories. The western settlement did not show any abandonment of areas and seems to have been stable until the catastrophe.

Comments are closed.